Bayer Filter obsolescence?

A brief article that might be of interest.

http://hosted.ap.org/dynamic/stories..._LIGHT_PHOTOS?
SITE=LAALE&SECTION=HOME&TEMPLATE=DEFAULT

I wonder if this would trade higher sensitivity for low color accuracy?

Eric Miller
www.dyesscreek.com

On Jun 14, 8:31 am, Eric Miller
wrote:
A brief article that might be of interest.

http://hosted.ap.org/dynamic/stories..._LIGHT_PHOTOS?
SITE=LAALE&SECTION=HOME&TEMPLATE=DEFAULT

I wonder if this would trade higher sensitivity for low color accuracy?

Eric Millerwww.dyesscreek.com

http://hosted.ap.org/dynamic/stories...LATE= DEFAULT


Seems a tad optimistic. I was under the impression that the
efficiency of current colour sensors was about 40-60%, compared with
90+% for some monochromatic sensors? In any case, sensitivity to
colours does not mean the thing can descriminate as to which colour
was which, I'd like to see how they do this, but I'm sure if it is a
major breakthrough, Kodak isn't really to reveal how unless they have
full patents. Thanks for the article.

"Eric Miller" wrote in message
...
A brief article that might be of interest.

http://hosted.ap.org/dynamic/stories..._LIGHT_PHOTOS?
SITE=LAALE&SECTION=HOME&TEMPLATE=DEFAULT

I wonder if this would trade higher sensitivity for low color accuracy?

Dunno.



http://www.dpreview.com/news/0706/kodaksens01.jpg

The sample here looks like crap! I'd rather use my 1D2N or a 5D at the same
ISO 1000.

On Jun 14, 8:31 am, Eric Miller
wrote:
A brief article that might be of interest.

http://hosted.ap.org/dynamic/stories..._LIGHT_PHOTOS?
SITE=LAALE&SECTION=HOME&TEMPLATE=DEFAULT

I wonder if this would trade higher sensitivity for low color accuracy?

Eric Millerwww.dyesscreek.com

Further,

http://www.dpreview.com/news/0706/07...sens.asp#press

"just bob" kilbyfan@aoldotcom wrote in
:

The sample here looks like crap! I'd rather use my 1D2N or a 5D at the
same ISO 1000.

I don't think the point is absolute noise; This may be a camera-phone-sized
sensor, and it clearly has unnecessary line noises. The point is the
difference between the two masks on the same sensor (unless they're
tricking us).



--


John P Sheehy

Eric Miller wrote in
:

I wonder if this would trade higher sensitivity for low color accuracy?

It would certainly cause lower color *resolution*, at the native pixel
frequency, but accuracy for a flat area of color should not be affected.

If you notice, the new pattern is also highly binnable (sp?). They may
intend optional hardware binning, to reduce read some of the read noise.

--


John P Sheehy

John Sheehy wrote:
(unless they're tricking us)

Why would a company who have already made some serious errors in the
digital business, and has been facing a hugely declining market for
their core business activities, resort to something as silly as trying
to trick us?

I actually doubt that this is a breakthrough anyway. Fuji's super CCD
uses luminance sensors combined with a bayer array which seems to work -
but not in an earth-shattering way.

Just thinking about it... what if they had a CMY filter instead of RGB? each
sensor element would then have twice the amount of light hitting it,
basically doubling it's sensitivity.
"Eric Miller" wrote in message
...
A brief article that might be of interest.

http://hosted.ap.org/dynamic/stories..._LIGHT_PHOTOS?
SITE=LAALE&SECTION=HOME&TEMPLATE=DEFAULT

I wonder if this would trade higher sensitivity for low color accuracy?

Eric Miller
www.dyesscreek.com

Doug Jewell wrote:
Just thinking about it... what if they had a CMY filter instead of RGB?

They couldn't patent it for prior art.
Wonder why they stopped building those ... probably something
to do with color accuracy ...

-Wolfgang

Doug Jewell wrote:
Just thinking about it... what if they had a CMY filter instead of RGB?
each sensor element would then have twice the amount of light hitting
it, basically doubling it's sensitivity.


Yes ... but to get RGB you have to do a subtraction, which
reintroduces the noise. In fact it is worse, because of
photon statistics.

(And yes, in fact, CMYK printing actually works on RGB because it is subtractive.)

The only real hope for better sensitivity is multiple sensors and
dichroic filters.

Doug McDonald